Testing WVO and biodiesel for water.

[Bill]

Julian your results at a quick glance look like they suffer the same problem as mine. The range of readings found in a sample overlap the range of readings found in another sample. This can make it unreliable as a method of measurement.
Ideally your RH results should be corrected for temperature and compared graphically or statistically.

[Julian]

I don't think they do.  I left the probe in the sample until it had stabilised for each reading and the first few show definite differences between wet and dry readings.

With the latter few some are spaced by long intervals and I suspect that the bio was absorbing moisture from the air making them read similar or same readings, but I'd need a definitive instrument to prove that.

Other things to bear in mind is that my sensor measures both RH and temperature.  On the cheaper probes there's a chip included in the sensor.  I'm not sure if that's just to digitise the signal of if it does any RH calculations.  If it doesn't then the Arduino library does the calculations.

The issue is that the RH is calculated for air.  I suspect the RH calculation for bio is very different.  On top of that we really want the measurement in PPM, as I believe it's an absolute measurement and not temperature dependent.

I've looked at a sensor from Switzerland that's designed to operate in oil, but that too gives readings in %RH.  I sat for half an hour in the local pub with a mate who has a PhD trying to understand the formula they issue for temperature correction and failed ... it just didn't seem to make sense!

So I'm at a bit of a loss at the moment!

[dgs]

Bio does absorb moisture from air a lot quicker than most people realise. I used to leave my finished bio in a 200 litre cone tank with a loose fitting top with bag filters in. On average the contents would increase by about 100ppm per week.

Still waiting for my meter, frustration,frustration.

[Bill]

Your max and min values are for wet 35 to 63 and dry from 26 to 63. That's overlap.
There is also an increase in readings as time progresses suggesting that there might be some sensor conditioning still going on as you were taking the measurements. I came across this with mine.
Try leaving the sensor in the dried sample for a couple of hours before taking any measurements and leave it in the dried bio between measurements but not necessarily powered up.
At the moment we are measuring samples at about dried+250ppm water and getting results which overlap. So are these sensors going to be sufficiently accurate to measure at the levels we need to find, which are ?.
Anyone any ideas about what should be in ?.

[Julian]

It would be good to know if a ppm reading can be taken directly from these sensors.  Obviously mine is a set-up arranged for air measurement, but it's just possible some of the other sensors can.

I might try asking the Swiss company, only problem is their sensors are £36 + VAT.

[Julian]

Your max and min values are for wet 35 to 63 and dry from 26 to 63. That's overlap.
There is also an increase in readings as time progresses suggesting that there might be some sensor conditioning still going on as you were taking the measurements. I came across this with mine.
Try leaving the sensor in the dried sample for a couple of hours before taking any measurements and leave it in the dried bio between measurements but not necessarily powered up.
At the moment we are measuring samples at about dried+250ppm water and getting results which overlap. So are these sensors going to be sufficiently accurate to measure at the levels we need to find, which are ?.
Anyone any ideas about what should be in ?.

I see what you mean, but don't you think the possibility of the samples absorbing water from the atmosphere could account for that?

The probe was put in the next sample immediately after taking the previous one so it was immersed for pretty much for three days.

[dgs]

I think they will be accurate enough Bill. For instance when i'm drying my bio I know the starting point after water washing is roughly 1200ppm after resting overnight. The turbo dryer will dry this to 50ppm in 13 hours, leaving the heat at around the 60deg mark. This is a little impractical, so what I want is a quick test to know when it gets down to that <300ppm level, without waiting a couple of hours for the s/b or setting up the KF.
With this method even if it was only accurate to (lets say) 100ppm then If I get a reading of 200ppm I would know the bio was dry enough. fair enough i wouldn't be able to say with any guarantee that it was 200ppm, but would be able to ensure that it would be no more than 300.

When I finally get the meter I will test various samples pf versus sandy brae and graph them.

[Julian]

Couple of points ...

... The turbo dryer ...

Seen you mention this a couple of times.  Is it a new fangled design not previously seen by humans before?

When I finally get the meter I will test various samples pf versus sandy brae and graph them.

I think that's probably the best starting point.  If very dry bio has a propensity for atmospheric moisture, then samples should really be double checked so we know the actual content at the time of the test.  Also we may be able to relate the readings you're getting from the sensor to ppm.

I think the Arduino can be programmed to work as a capacitance meter, I'll see what I can find out.  If that's possible then I'd think the readings could be manipulated in the firmware to give a ppm reading.

 

[Julian]

... I think the Arduino can be programmed to work as a capacitance meter, ...

Yes it can, Julian.

I found this page ... http://wordpress.codewrite.co.uk/pic/2014/01/21/cap-meter-with-arduino-uno/

I won't pretend to understand the theory, but I can probably have a bash at getting it to work.

[Julian]

Seems to work reasonably well.  This is the result of sticking a 4.7µf capacitor in pins A0 and A2 ...




It took longer to get the com port working than anything else!

I might have a look for other programs.

[oakwoodtv]

The readout is in Picofarads and the capacitor is 4.7µf is that a typo.

[Julian]

I've got confused with my micro and picos haven't I?

It should be 4700000 pF.

Hmm, I'll have another look!

[Julian]

The sketch above should have been calibrated, but I can't see that it was that far out.  Another issue could have been I was using a different Arduino board that may have had a different capacitance ... don't know it's all a bit beyond me.  I just plug things in the way they say without knowing what I'm doing!

I found another sketch that still only uses two pins, doesn't appear to need calibrating and changes denomination (if that's the right word) automatically.  I used a Uno board this time in case it makes a difference.

It's here ... http://www.circuitbasics.com/how-to-make-an-arduino-capacitance-meter/     last one on the page.

This is the readout using a 1µf capacitor (hope the nanos and micros are right this time!) ...





Not quite sure what the stuff on the right in brackets is.  I think it's something to do with the time to charge the capacitor, so probably of no use to us and can easily be omitted in the sketch.

[Julian]

Having apparently got a capacitance measurement working, I thought I'd dismantle one of the cheap DHT11 sensors I bought.

Cutting the plastic cage off, reveals a tiny PCB with what looks like 5 surface mount resistors (but may be other components) and a 14 pin, unmarked chip on one side and the capacitance sensor on the other side.

I un-soldered the sensor, re-soldered it to a couple of wires and pins and plugged it into the Arduino board thus ...




"Herrrrrring" on the sensor gives the following readings ...





Quite a large range, 24pF to 2µF to play with.  Not sure what happened with the readings in the middle,  but it's been very stable sitting on the bench for the last ½ hour.

[Julian]

Now available as a portable version with LCD display.